Attribution of climate change and human activities to streamflow variations with a posterior distribution of hydrological simulations

: Hydrological simulations are a main method of quantifying the contribution rate (CR) of climate change (CC) and human activities (HAs) to watershed streamflow changes. However, the uncertainty of hydrological simulations is rarely considered in current research. To fill this research gap, based on the Soil and Water Assessment Tool (SWAT) model, in this study, we propose a new framework to quantify the contribution rateCR of climate changeCC and human activitieHAs based on the posterior histogram distribution of hydrological simulations. In our new quantitative framework, the uncertainty of 15 hydrological simulations is first considered to avoid quantifyreduce the impact of the phenomenon of "equifinality for different parameters", which is common in hydrological simulations. The Lancang River (LR) Basin in China, which has been greatly affected by human activitieHAs in the past two decades, is then selected as the study area. The global gridded monthly sectoral water use data set (GMSWU), coupled with the dead capacity data of the large reservoirs within the LR basin and the Budyko hypothesis framework, are used to compare the calculation result of the novel framework. The results show that (1) the annual 20 streamflow at Yunjinghong station in the Lancang River Basin changed abruptly in 2005, which was mainly due to the construction of the Xiaowan hydropower

4 of hydropower stations have been constructed in the LR Basin to meet the flood control and drought relief requirements of downstream countries and the power needs of Southwest China. Therefore, it is particularly important to quantify the CRcontribution rate of climate changeCC and HAhuman activities to streamflow changes in the LR Basin. However, so far, there are still few corresponding studies. Han et al. (2019) chose the Lancang RiverLR Basin (LRB) as the study area and then divided the research period into three periods, the natural period, transition period, and impacted period, and combined them with the 115 construction time of six large hydropower stations in the LR area. Finally, they found that the CRcontribution rate of HAhuman activities during the impact period exceeded 95%, using the coupled routing and excess storage (CREST) model, which was probably due to the construction of the Nuozhadu hydropower station. However, there are still areas for improvement in their research: 1) the results of the hydrological simulation were relatively poor (with monthly NSE = 0.57 for the whole study period), and 2) the uncertainty involved in hydrological simulations was not considered. 120 In this paper, the breakpoint of the change in flow regimes was identified using the Mann-Kendall break point test. Then, the study period was divided into a natural period (before the breakpoint) and an impacted period (after the breakpoint). The Soil and Water Assessment Tool (SWAT) model was used for monthly streamflow simulation at the Yunjinghong station. Next, the monthly SWAT model was calibrated and validated using the sequential uncertainty fitting procedure version 2 (SUFI-2) (Abbaspour et al., 2004). Uncertainty analysis was also conducted with the SUFI-2 method, and then the posterior histogram 125 frequency distribution (HFD) of the CR of CC and HAs was obtained. Finally, the proposed quantification framework was compared with two other methods: one was the Budyko framework, and the other was to use the LR Basin's gridded monthly sectoral water withdrawals in the period from 1971 to 2010 (Huang et al., 2018) together with the dead reservoir storage capacity data of the six constructed hydropower stations along the main stream of LR, to separate the CRcontribution rate of HAhuman activities. 130

Study area
The Lancang River (LR) originates in the northeastern Tanggula Mountains, Qinghai Province, China, and flows through China's Qinghai Province, Tibet Autonomous Region and Yunnan Province. It is the largest international river in Southeast Asia, and it is called the Mekong River after it flows out of China. Its main stream has a total length of ~2161 km and a total catchment 135 area of ~160000 km 2 (Han et al., 2019;Li et al., 2017a). The topography of the LR is characterized by high northern and low southern portions; the maximum elevation in the northern mountainous area can reach ~5871 meters, while the lowest elevation in the downstream area is only ~547 meters (Fig. 1). This steep terrain difference also leads to the LR having a large potential for hydropower resources. During the past few decades, Huaneng Lancangjiang Hydropower Co., Ltd. constructed six large hydropower stations (i.e., Gongguoqiao, Xiaowan, Manwan, Dachaoshan, Nuozhadu and Jinghong) on the main stream of the LR 140 to meet the demands for power and irrigation water in Southwest China ( Fig. 1 and Table 1) (Han et al., 2019;Hennig et al., 2013;Xue et al., 2011). At the same time, the construction of these hydropower stations has greatly reduced the risk of flooding in downstream countries and brought great convenience to using water for downstream agricultural irrigation. Detailed information on the six constructed hydropower stations is outlined in Table 1. These data are mainly collected from https://opendevelopmentmekong.net/topics/hydropower/, as well as from other published related literature (Han et al., 2019;Xue 145 et al., 2011;Hennig et al., 2013;Tilt and Gerkey, 2016).

5
The LR features an arid climate in the upper mountainous areas, while the lower reaches are dominated by humid climates. The average annual precipitation of the whole basin is ~ 870 mm based on a 55-year record (from 1961 to 2015) using the China Gauge-based Daily Precipitation Analysis (CGDPA) (Xie et al., 2007;Tang et al., 2019). Due to the influence of the westerlies and the Indian Ocean monsoon, the precipitation in the LR has obvious seasonal changes, and the precipitation from June to 150 September accounts for more than 70% of the annual precipitation (Jacobs, 2002). Correspondingly, the streamflow of the LR also shows seasonality, and the floods are mostly concentrated from June to September.  (Notation: Dead storage capacity refers to the storage capacity below the dead water level of the reservoir, which does not participate in runoff regulation during the normal operation of the reservoir.) 160

Data sets
The China Gauge-based Daily Precipitation Analysis (CGDPA) product was developed by the China Meteorological Administration (CMA) using data from ~2400 ground-based national weather stations across China Xie et al., 2007;Shen et al., 2014). It provides daily precipitation, maximum temperature, minimum temperature, relative humidity, and wind speed data at a 0.25-degree spatial resolution from 1961 to 2015 (http://cdc.nmic.cn). Previous studies have successfully 165 applied this product to multiple research areas in China Han et al., 2019). The daily streamflow data from Yunjinghong station for the time period from 1961 to 2015 were collected from the Information Center of the Ministry of Water Resources and the local water resources management department. The digital elevation model (DEM) used in this study was downloaded from NASA's Shuttle Radar Topography Mission (SRTM) data bank at a spatial resolution of ~90 meters (http://srtm.csi.cgiar.org/), which was used to generate the watershed 170 boundary, slope and sub-watershed data in the SWAT model (Arnold et al., 2012a). The Harmonized World Soil Database (version 1.2) (HWSD v1.2) at a spatial resolution of ~1 km was downloaded from the Food and Agriculture Organization of the United Nations, and this data set contains two layers of soil. The land use and cover data with a spatial resolution of ~1km were collected from the Geospatial Data Cloud (http://www.gscloud.cn/). In this study, to analyze the land use change in the LR during the historical period, we collected five periods of land use data in the 1980s, 1990s, 2000s, and from 2010 to 2015, and this data set 175 was downloaded from the Geographic Information Monitoring Cloud Platform (http://www.dsac.cn/), with a spatial resolution of 30 meters.
The global gridded monthly sectoral water use (GMSWU) data set for 1971-2010 was obtained from https://zenodo.org/record/1209296#.XsJmiTNlsSJ. This data set was developed by Huang et al. (2018), and it provides the global domestic water use, irrigation water use, livestock water use, manufacturing water use and mining water use with a spatial 180 resolution of 0.5 degrees. This dataset is used in this study because it is difficult to collect water withdrawal data related to HAs in the LR Basin, and this dataset has been successfully applied in this basin in other studies (Han et al., 2019). And wWe used this data set here to roughly separate the effects of HAhuman activities in the LR. For more technical information about this set of products, the readers can refer to Huang et al. (2018) and Han et al. (2019). Furthermore, detailed information on six large dams in the main stream of the LR was collected from Open Development Mekong 185 8 (https://opendevelopmentmekong.net/topics/hydropower/) and Huaneng Lancang River Hydropower Inc. It mainly includes the dates when the rivers start to be closed, when these dams were fully put into use, their dead storage capacity, their total storage capacity, and other information.

The novel proposed framework 190
Hydrological simulation is one of the main methodologies to quantify the CRcontribution rate of climate changeCC and human activitieHAs to streamflow variations; however, in the past, related studies have rarely considered the uncertainty involved in hydrological simulations (Farsi and Mahjouri, 2019). In this section, we will introduce a new quantitative framework to avoid quantifyreduce the influence of thethe common phenomenon of "equifinality for different parameters" in hydrological simulation on the quantitative results, by constructing the posterior distribution of streamflow simulations during the implementation process. 195 The specific implementation flowchart is shown in Fig. 2. The detailed execution steps are shown as follows.
Step 1: Inspection of break points in the annual streamflow sequence; based on the result of break point test, the entire time series is divided into a natural period (before the break point) and an impacted period (after the break point).
Step 2: Sensitivity analysis of the parameters in the hydrological model.
Step 3: According to the results of the parameter sensitivity analysis, selection of the more sensitive parameters and input of 200 the hydrometeorological data of the natural period to calibrate the hydrological model with 1000 runs.
Step 4: Selection of the parameter sets with Nash-Sutcliffe efficiency coefficients (NSE) is greater than 0.75 in 1000 simulations, input of the hydrometeorological data of the impacted period, and further calculation of the CR of CC and HAs to the streamflow change corresponding to each simulation result.
Step 5: Construction of the posterior histogram distribution (PHD) of the CR of CC and HAs (with a 5% step), and then the 205 histogram with the highest frequency is treated as the uncertainty CRcontribution rate interval of CC and HAs to the streamflow change.
Step 6: The arithmetic mean of the results in the interval is treated as its true CRcontribution rate.
In step 4, to ensure the number of streamflow simulation samples, we set the simulation results with NSE is greater than 0.75 to at least 500 times. If the setting is not met, then step 3 is repeated until the cumulative simulation times are greater than 500 210 times.

Mann-Kendall test
In this step, the trends and break points of the hydrometeorological data are detected using the nonparametric  respectively. The main consideration of using the Mann-Kendall test is that this method assumes no particular distribution for the tested time series (Song et al., 2019;Xu et al., 2018). Significance levels of α = 0.01 and 0.05 is used in this study.

Mann-Kendall monotonic trend test
The Mann-Kendall (MK) monotonic trend test was developed by Mann (1945), Kendall (1975) and Gilbert (1987), which 220 has been widely used to detect the presence of an upward or downward trend of the hydrometeorological time series, and the advantage of this test is that the time series does not need to follow a certain distribution (Hamed and Ramachandra Rao, 1998).
This method first tests whether to reject the null hypothesis ( 0 H : no monotonic trend) and accept the alternative hypothesis ( a H : with monotonic trend) for a significance level of  . The defined statistic S can be calculated by the following equation: where k x is the data in the order over time, x 1 , x 2 , …, x n-1 , which means the time series obtained at times 1, 2, …, n-1, respectively; j x is another time series over time x k+1 , x k+2 , …, x n ; n is the length of the data set record; and is a sign function that takes on the values of 1, 0, or -1 based on the sign of jk xx − , and its values can be calculated by the following equation: After calculating the S sequence, the variance of S can be computed as follows: where n is the length of the time series; g is the length of any given tied group and p t is the length of the data set series in the th p group. Then, the defined test statistic MK Z can be transformed from the statistical value S , and the equation is as follows:

Mann-Kendall break point test
The break point of the hydrometeorological time series denotes a change from one stable state to another stable state (Xu et al., 2018). It occurs when the climate system breaks through a certain threshold. The Mann-Kendall break point test has been widely 240 used to test break points for hydrometeorological time series, signaling when abrupt changes start (Sneyers, 1991). This test method is used to determine the break point of the observed annual streamflow in this study. The defined statistic k UF is obtained by the following formulas: where i x is the variable to be tested and n is the total number of data points. The expectation After identification of the break points in the annual streamflow series, the study period is divided into a "natural period" (before the break point) and an "impacted period" (after the break point) (Wang et al., 2015;Bao et al., 2012). The "natural period" means that there is no significant increase or decrease in streamflow during this period, and it also means that relatively slow climate changeCC is the dominant factor, and that the impact of HAhuman activities is very small during this period. Consequently, 260 the impacted period indicates to a significant change in streamflow during this period, mostly due to factors such as the construction of water conservancy engineering facilities, increased water consumption for irrigation, changes in land use and increased water consumption in cities and towns.

SWAT model
The Soil & Water Assessment Tool (SWAT) model is a semi-distributed, physical process-based hydrological model developed 265 by the Agricultural Research Service of the United States Department of Agriculture (USDA-ARS) (Arnold et al., 1998). The SWAT model first divides the study area into several subbasins based on DEM data, and then each subbasin is further divided into several HRUs (Hydrologic Response Units) based on land use and soil data sets. Then, streamflow generation at the subbasin scale is calculated following the principles of water balance and energy balance after inputting the meteorological data sets. Finally, the total flow of river basin exports is calculated according to the Muskingum method Arnold et al., 2012b). 270 We chose to use the SWAT model in this study because numerous published studies have proven that this model has excellent performance in hydrological simulations across the world Zhao et al., 2018a;Zhao et al., 2018b;Lee et al., 2018). The calibration of model parameters is executed using the independent software SWAT-CUP, which was developed by Abbaspour et al. (2007). This software is freely available and provides five parameter calibration and uncertainty analysis methods. 275 In this study, the sequential uncertainty domain parameter fitting version 2 (SUFI-2) algorithm (Abbaspour et al., 1997;Abbaspour et al., 2004) was used to perform parameter calibration and uncertainty analysis, because this method has proven to have the advantages of shorter calculation time, ease of implementation and ability to set arbitrary objective functions (Zhao et al., 2018a;Tuo et al., 2016;Wu and Chen, 2015). The performance of the SWAT model was evaluated by the Nash-Sutcliffe efficiency coefficient (NSE) (Nash and Sutcliffe, 1970) and relative error (RE): 280 where , and , are the observed and simulated streamflow, respectively; ̅̅̅̅̅̅ is the mean value of the observed streamflow; is the total number of days or months in the calibration period; and and are the mean annual simulated and observed streamflow, respectively. 285

Construction of the Posterior Histogram Distribution of contribution rate
In this section, we introduce how to calculate the contribution rate (CR) of climate change (CC) and human activities (HAs) to streamflow variations and how to construct the posterior histogram distribution (PHD) of the CR to consider the uncertainty of hydrological simulations.

Contribution rateCR of climate changeCC and human activitieHAs 290
A schematic diagram of the attribution evaluation of streamflow changes is shown in Fig, 3. ∆Q in the figure represents the amount of change in the observed streamflow during the impacted period based on the natural period, while ∆ and ∆ ℎ represent the amount of streamflow change caused by climate changeCC and human activitieHAs, respectively. The total change in the annual streamflow can be calculated using the following formula: where ̅̅̅̅ and ̅̅̅̅̅ are the mean annual observed streamflow (m 3 /s) in the impacted period and natural period, respectively. The hydrological and meteorological data in the natural period are input into the SWAT model, and using the SUFI-2 method to calibrate the model, a set of parameters represents the characteristics of catchment under natural conditions with less impact from HAhuman activities. Then, this set of parameters is brought back into the SWAT model using the meteorological data of the impacted period. Based on the above simulation results, the climate changeCC induced in streamflow can be calculated as follows: 305 where ̅̅̅̅ and ̅̅̅̅̅ represent the mean simulated annual streamflow (m 3 /s) for the impacted period and natural period, respectively.
Thus, the streamflow change induced by HAhuman activities can be calculated by the following equation: After the calculation of ∆ and ∆ ℎ , the CRcontribution rate of climate changeCC and HAhuman activities to streamflow 310 changes, which are defined as and ℎ , respectively, can be estimated as: Equations 12 to 15 are also applicable to quantify the CRcontribution rate of climate changeCC and human activitieHAs to streamflow changes on a monthly scale. 315

Construction of the PHD of the CR of CC and HAs
Before the construction of the PHD of the CR of CC and HAs, the sensitivity of the parameters of the SWAT model is first conducted. Based on the related published literature (Zhao et al., 2018a;Yang et al., 2008;Malagò et al., 2015) and the authors' experience, the Latin-Hypercube and global sensitivity methods were used to perform the uncertainty analysis (Abbaspour et al., 2007). The global sensitivity analysis method is the estimation of the average change in the objective function caused by the 320 change in each parameter, and all parameters change during the whole process. A t-test was used to identify the relative sensitivity 14 of each parameter. Considering the influence of the snowmelt streamflow process upstream of the LR Basin on the hydrological simulation, 22 parameters were selected and the details of these selected parameters are shown in Table 2. According to the suggestion of Abbaspour et al. (2004), 500 simulations were set up to implement the sensitivity analysis. The t-stat and P-values were used to measure which parameters were more sensitive, where a larger absolute t-stat value and a smaller absolute P-value 325 represent a higher sensitivity of a given parameter. Based on the sensitivity analysis results, 9 parameters with the highest sensitivity were selected to re-calibrate the model with 330 1000 simulations. According to the recommendations in Tuo et al. (2016) and Moriasi et al. (2007), the performance of the hydrological simulation can be divided into four grades based on the NSE values: very good performance (0.75 ≤ NSE < 1), good performance (0.65 ≤ NSE < 0.75), satisfactory performance (0.5 ≤ NSE <0.65) and unsatisfactory performance (NSE < 0.5).
According to this evaluation standard, we selected simulation results with NSE greater than 0.75 out of 1000 simulation results to construct the posterior histogram frequency distribution (PHD) of the contribution rate CR of climate changeCC and human 335 activitieHAs to streamflow changes using the method introduced in section 3.4.1. Note that to reduce the random error caused by the number of samples, we set the number of simulations with NSEs ≥ 0.75 to be more than 500; that is, we needed to repeatedly use Latin hypercube sampling and the SUFI-2 algorithm until the number of simulation results that met the conditions was more than 500. Then, the CR of more than 500 groups of CC and HAs to streamflow change was calculated. Finally, the posterior histogram distribution (PHD) of the CR of CC and HAs was constructed in 5% steps. At this stage, the histogram column with 340 the highest frequency in the PHD was selected as the result of quantitative analysis, which considered the uncertainty, and the arithmetic average of all results in the column was used as the actual value of the CR of climate changeCC and HAhuman activities.

Comparison of the newly developed quantification method with other two methods
In order to evaluate the calculation accuracy of the novel framework proposed in this study to quantify the CR of CC and HAs to streamflow changes, the Budyko framework was used first. This framework was developed by Budyko (1961) and links climate 345 variability to streamflow (Q) and actual evapotranspiration (AE) through the assumption that the long-term average annual catchment AE is determined by the catchment average precipitation (P) and the catchment potential evapotranspiration (PET) (Liu and Liang, 2015). Over the past few decades, the Budyko framework and its variants have been widely used to conduct climate changeCC and human activityHA attribution analyses of streamflow changes (Liu et al., 2017;Han et al., 2019;Xin et al., 2019).
According to its theoretical assumptions, the multiyear average water balance within the catchment can be expressed as follows: 350 where P, Q, and AE represent the multiyear average precipitation (mm), streamflow (mm) and actual evapotranspiration (mm), respectively; ∆S (mm) is the change in the amount of water storage at the watershed scale, and it is reasonable to assume that it is equal to 0 on the multiyear average scale. According to Zhang et al. (2001), the AE can be calculated by the following formula: where is the plant-available water coefficient which is related to the vegetation type of the catchment.; According to the method for selecting the value of provided in Zhang's research (Zhang et al., 2001), and based on the multi-year average AE/P (0.55) and PET/P (0.96) values in the LR Basin, this study set the value of to 0.5.it is set to 0.5 in this study.
The changes in the catchment streamflow due to climate changeCC, which are mainly characterized by precipitation (P) and actual evapotranspiration (AE), can be expressed as follows: 360 where ∆ (mm) represents the streamflow changes induced by climate changeCC; and represent the sensitivity of streamflow to precipitation and actual evapotranspiration, respectively; and ∆ and ∆ are the changes in precipitation and actual evapotranspiration in the impacted period compared to the natural period, respectively. The sensitivity coefficients and are defined as follows: 365 where DI is the dryness index which is equal to ⁄ .
Through the above formulas, we can separate the CRcontribution rate of climate changeCC to streamflow variations, and further compare it with the calculation results of the new method proposed in this paper. 370 In addition to the Budyko framework, we also used the GMSWU data introduced in Section 2.2 and the reservoir dead storage capacity data to roughly separate the CR of HAs from the streamflow changes in the LR Basin. The GMSWU data set provides five types of water withdrawals (i.e., irrigation, livestock, domestic use, mining, and manufacturing) within the period of 1970 to 2010 in the LR Basin, and it was generated by downscaling country-scale estimates of different sectoral water withdrawals from the Food and Agriculture Organization (FAO) of the United Nations AQUASTAT, which ensured its good accuracy (Huang et 375 al., 2018). Here, AQUASTAT refers to FAO's Global Information System on Water and Agriculture (http://www.fao.org/aquastat/en/). Catchment-scale annual water use data were calculated by spatially averaging all grids within the LR Basin, and then streamflow changes caused by each type of water use were obtained using the average annual water use value during the impacted period minus that during the natural period. As shown by Han et al. (2019) and Zhao et al. (2012), during the past two decades, dam construction has been the most significant human activityHA affecting the streamflow changes 380 in the LR Basin. Therefore, in this study, we converted the dead storage capacity of 6 large reservoirs ( Table 1) into units of millimeters according to their watershed control area because the impact of the reservoir on the outlet flow of the watershed can be used as its minimum impact value on the multiyear average scale. Note that the CR of CC and HAs calculated by the above two methods was not the actual true value but rather an estimate. It should be pointed out that here we use two seemingly simpler methods to verify the computational results of the new framework proposed in this study. However, this does not reduce the 385 innovation of this study, as the new framework has the following significant advantages over the other two methods: 1) The new framework can perform quantitative calculations on the annual and monthly scales; 2) It has relatively less data requirements; 3) It has a more explicit physical meaning. We used these two methods to compare with the newly proposed framework developed in this study. In addition, although the calculations of these two methods are simple, they have the following shortcomings compared with the CR of CC and HAs derived from hydrological simulation: 1) the CR of CC and HAs to streamflow variations 390 can only be calculated on a multiyear average scale and cannot be calculated on a monthly or seasonal scale; 2) they have higher requirements for data input (e. g., data related to reservoirs); and 3) they have relatively less physical meaning compared with the streamflow simulation of the distributed hydrological model.  during the year was mainly caused by the operation of reservoirs within the basin, because reservoirs often release flows during 415 dry periods (from January to May) to alleviate possible droughts in the downstream areas, and they store water during wet periods (from June to October) to reduce the flood control pressure in the downstream area below the reservoir.

Trends and break points of the mean areal precipitation and temperature
The time series and MK break point test results of the annual areal precipitation and mean temperature for the LR Basin from 1961 to 2015 are presented in Fig. 6. In general, changes in the annual precipitation were more complicated than changes in the mean temperature in the LR Basin. The precipitation showed a fluctuating trend, while the mean temperature almost showed a continuous rising trend throughout the study period. 425 As shown by the time series of the annual precipitation in the LR Basin in Fig. 6 (a), there was a slightly decreasing trend in the LRB during the last 55 years, especially in the past 10 years, but this trend was not  The time series of the annual mean temperature in the LR Basin presented in Fig. 6 (b) shows that the annual mean temperature in the study area changed relatively smoothly before 1998. After 1998, the temperature began to rise significantly and exceeded the significance level of 0.01. The annual mean temperature in 1963 reached 5.2 ℃, the coldest temperature in the study period. The hottest year was 2009, during which the mean temperature was 7.2 ℃. In terms of changes in the UF value, the 440 mean temperature showed a fluctuating trend from 1961 to 1968, and then continued to rise until it exceeded the significance level of 0.05 in 1991 and exceeded the significance level of 0.01 in 1998. The break point of the annual mean temperature was detected in 1997.
The MK monotonic trend test statistics of the annual and monthly precipitation and mean temperature for the LR Basin from 1961 to 2015 are presented in Fig. 7. The annual precipitation in the study area showed an insignificant decreasing trend (Z statistic 445 = -0.55), while the annual average temperature showed a significant increasing trend (Z statistic = 6.02) and exceeded the significance level of 0.01. The monthly change in precipitation also showed a fluctuating trend. The increasing trend of precipitation in April and the decreasing trend of precipitation in June exceeded the significance level of 0.05, while the trends of precipitation in other months were not significant (|z statistic| < 1.96). The trend of the monthly mean temperature was relatively simple. Except for the increase in the mean temperature in November, which passed the significance level of 0.05, the increasing 450 trend of the mean temperature in all other months passed the significance level test of 0.01. This also means that the climate in the study area has been gradually warming and drying during the past 55 years.

Sensitivity analysis of the SWAT model parameters
As descripted in Section 3.4.2, the sensitivity of 22 selected parameters was evaluated using the SWAT-CUP software (Abbaspour et al., 2007;Abbaspour et al., 1997), and this software integrates the global sensitivity analysis method and the parameter optimization methods (such as SUFI-2). The SWAT-CUP can using the SUFI-2 and global sensitivity analysis methods. 460 SUFI-2 performs a combined optimization and uncertainty analysis using a global search procedure and can deal with a large 20 number of parameters through Latin hypercube sampling. The sensitivity evaluation indexes, the t-Stat and P-value, of 22 parameters are shown in Table 3. Obviously, the sensitivity ranks of the parameters calculated based on SUFI-2 showed that ALPHA_BNK has the highest sensitivity, followed by CH_K2, SOL_BD, GW_REVAP, SFTMP, CN2, SOL_K, SMTMP and ALPHA_BF, whereas the other 14 parameters have less sensitivity for the streamflow simulation. ALPHA_BNK mainly controls 465 the baseflow process within the watershed, and this parameter has also proven to have high sensitivity in other relevant studies (Wu and Chen, 2015), especially in mountainous areas. CH_K2 and ALPHA_BF are mainly related to groundwater runoff, CN2 is the SCS runoff curve number, and these parameters all have higher sensitivity in many published articles on the SWAT model parameter sensitivity (Zhao et al., 2018a;Wu and Chen, 2015). Other parameters with high sensitivity, such as SFTMP and SMTMP, which mainly control the snowmelt process in the basin, also indicate that snowmelt runoff plays an important role in 470 the recharge of the LR Basin . Based on the above sensitivity analysis results, the top 9 parameters of the sensitivity ranking were selected for further research.

Results of the SWAT simulations
As mentioned above, the 9 parameters with the highest sensitivity rankings that controlled different stages of the basin's streamflow production and flow concentration were selected to re-calibrate the model using the SUFI-2 method, and the number of simulations was set to 2000. To reduce the influence of the initial value of the model parameters on the simulation results, during the model parameter calibration process, 1961 and 1962 were set as the warming-up period. Table 4 shows the evaluation 480 metrics of the simulation using the SWAT model at a monthly scale with the largest NS value. For the calibration period from 1963 to 1990, the NSE and RE were found to be equal to 0.94 and -10.62%, respectively; for the validation period from 1991 to 2004, the model performance was slightly better than that in the calibration period, and the NSE and RE were 0.95 and -8.65%, respectively. For the whole period from 1963 to 2004, the NSE (0.94) and RE (-9.97%) were also satisfactory. According to the requirements of the Information Center of the Ministry of Water Resources, the data provider, this study standardized the observed 485 and simulated runoff curves of the Yunjinghong station. Fig. 8 shows the normalized monthly observed and simulated streamflow 21 at Yunjinghong station from 1963 to 2004 and the histogram of the mean monthly precipitation in the LR Basin. As seen from Fig. 8 (a) and Fig. 8 (b), the SWAT model can simulate the flow processes very well and almost perfectly match the observed streamflow curve. Note that the simulated streamflow overestimated the floods in individual years (1973, 1985 and 1995 in Fig.8 (b)), which might be caused by the uncertainty of the precipitation product (Han et al., 2019). In summary, the SWAT model can 490 better simulate the streamflow process at Yunjinghong station on a monthly scale; therefore, this model is considered suitable for the next part of the research. According to the method described in Section 3.4.2, simulations with NSEs greater than 0.75 among the 1000 simulations were selected. Fig. 9 shows the number of simulations with 0.75 ≤ NSE < 0.8, 0.8 ≤ NSE < 0.85, 0.85 ≤ NSE < 0.9 and 0.9 ≤ NSE < 500 0.95 during the calibration period (1963 -1990), the validation period (1991 -2004) and the whole period (1963 -2004) on a monthly scale. In summary, there were 575 simulations with NSEs greater than 0.75 out of 1000 simulation results during the calibration period, the validation period, and the whole period. Clearly, the NSEs of most simulation results were between 0.75 22 and 0.9, with 533, 537 and 533 simulations in the calibration period, the validation period, and the whole period, respectively, and only a few simulation results had NSEs greater than 0.9. In the different periods, the model performed well in the validation period 505 compared with that in the calibration period, which indicated that the model has good predictive ability in the LR Basin. Fig. 9. Number of simulations with NSEs greater than 0.75 during the calibration (1963 -1990), validation (1991 -2004), and whole periods (1963 -2004).

Quantification of the impacts considering the uncertainties at the annual scale
The 575 simulations with NSEs greater than 0.75 were selected to construct the posterior histogram frequency distribution (PHD) of the CRcontribution rate of climate changeCC and HAhuman activities to streamflow changes in the LR Basin. Fig. 10 shows the number of simulations of the climate changeCC CRcontribution rate in 5% intervals and their corresponding NS box plots. In total, 167 out of 575 simulations calculated that the CRcontribution rate of climate changeCC in the LR Basin to runoff 515 reduction was 40% -45%, and the average NSE was 0.84. Then, 131 and 92 of the simulation results had calculated climate CRcontribution rates of 35 -40% and 45 -50%, respectively. The CRcontribution rate of climate changeCC in other intervals had relatively few simulations. The NSE value of the climate changeCC CRcontribution rate between 70-75% was the largest (NSE = 0.86), but it had only 1 simulation. Therefore, when using hydrological simulations to quantify the CRcontribution rate of climate changeCC and HAhuman activities to the streamflow change of the watershed, not only the merits of the model 520 performance but also the uncertainty of the model simulation should be considered. In general, according to the results calculated 23 by the new quantitative framework proposed in this paper, streamflow changes in the LR Basin duo to climate changeCC accounted for 40-45% (with an average CRcontribution rate of 42.6%), and the corresponding HAhuman activities accounted for 55-60% (with an average CRcontribution rate of 57.4%). 525 Fig. 10. Histogram of the number of simulations of the CRcontribution rate (with 5% steps) of climate change to streamflow reduction in the LRLancang River Basin at the annual scale and corresponding Nash-Sutcliffe Efficiency box plots. Table 5 shows the average values of the main hydrological and meteorological elements and their changes during the natural period and the impacted period. During the impacted period, compared with the natural period, the multiyear average streamflow decreased by 396 m 3 /s (86.5 mm), the precipitation decreased by 25 mm; as basin wide temperatures increased, the mean potential 530 evapotranspiration and temperature in the basin increased by 6.4 mm and 0.9°C. In terms of relative changes, the streamflow decreased by 22%, but precipitation and potential evapotranspiration changed by -2.9% and 6.4%, respectively, which may indicate that the streamflow reduction in the LR Basin was mainly caused by HAhuman activities. (Notation: PET = "Potential evapotranspiration" and T = "Temperature")

Quantification of the impacts considering the uncertainties on a monthly scale
The monthly CRcontribution rate of climate changeCC and HAhuman activities to the changing streamflow at Yunjinghong station was also analyzed using the new framework proposed in this study, and the results are shown in Fig. 11. In general, only June and November had a large CRcontribution rate of climate changeCC, which reached 95 -99.9% and 70 -75%, respectively, 540 while the CRcontribution rate of climate changeCC in the other 10 months was relatively small. The trends of the streamflow and the precipitation and mean temperature in the study area shown in Fig. 5 and Fig. 7 indicate that the streamflow in June and November showed a decreasing trend (Fig. 5), while the precipitation in June decreased significantly (passing the significance level of 0.05), and the temperature increased significantly (passing the significance level of 0.05) (Fig. 7). This significant decrease in precipitation and the significant increase in temperature were the main reasons for the decrease in the streamflow in June; that 545 is, the decrease in the streamflow in June was mainly caused by climate changeCC. The main factors that led to the decrease in the streamflow in November were also the decrease in precipitation and the significant increase in temperature (Fig. 7). From the results of each month, the CRcontribution rate of climate changeCC in March and April was the smallest, reaching 10 -15%; followed by July (15 -20%); May, August, and September (20 -25%); October (25 -30%); January and February (30 -35%); and December (45 -50%). 550 Fig. 11. Histogram of the number of simulations of the contribution rate (CR) (with 5% step) of climate change to streamflow reduction in the LRLancang River Basin on a monthly scale.
The mean CRcontribution rate of climate changeCC and HAhuman activities at the monthly scale, which was calculated by averaging the CRcontribution rates of all simulation results within the highest frequency, is displayed in Fig. 12 (left panel), and 555 the monthly precipitation, potential evapotranspiration and runoff depth during the natural period and the impacted period are shown in the right panel of Fig. 12. Overall, the monthly CRcontribution rate was consistent with the annual results, and the 25 CRcontribution rate in a total of 10 months was mainly due to HAhuman activities that led to a decrease in the streamflow in the LR Basin. It is worth noting that the CRcontribution rate of climate changeCC in June reached 96%. The panel in the right of Fig.   12 shows that the precipitation in June during the impacted period was significantly reduced compared with the natural period 560 (with a 20.2 mm decrease). At the same time, the increase in potential evapotranspiration in June was also relatively obvious (with a 9.2 mm increase). Fig. 12 (right panel) clearly shows that the streamflow in the LR Basin during the impacted period was significantly reduced compared with the natural period in June to October, and the precipitation had little change, except in June.
Therefore, we can conclude that the main reason for the decrease in the streamflow in the LR Basin was HAhuman activities, as shown in the left panel. In this study area, the main cause of the streamflow changes was mainly due to the construction of 565 reservoirs (such as Manwan and Xiaowan), and at the same time, the water storage of these water conservancy facilities during the flood period also provides engineering support for protecting the safety of downstream life and property. Conversely, during the dry season (from January to May), the streamflow in the impacted period showed an increasing trend compared with the natural period, and the increase in runoff during these five months was mainly due to HAhuman activities (Fig. 12, left panel), which might have been caused by the release of water from the reservoirs during the dry season. For example, in 2016, due to the 570 influence of El Niño, the countries along the lower Mekong River all suffered severe drought. The Chinese government immediately asked the Jinghong Reservoir to release water urgently, which effectively helped downstream countries mitigate a series of possible effects caused by drought and water shortages (Li et al., 2017b).

Comparison with the other two methods
In this sub-section, the new proposed framework that considers the uncertainty of hydrological simulations was compared with 580 the Budyko framework, five sections of water withdrawal data from the LR Basin and the equivalent streamflow depth converted from the dead storage capacity of six large hydropower stations. Table 6 shows the CRcontribution rate of climate changeCC and HAhuman activities to annual streamflow changes at Yunjinghong station, which was calculated from the Budyko framework. The actual evapotranspiration was calculated from the annual precipitation minus the annual streamflow depth. As shown in Table 6, compared with the natural period, the precipitation 585 and streamflow depth in the impacted period showed a decreasing trend. The precipitation decreased by 25 mm and the streamflow depth decreased by 86.5 mm. In contrast, the actual evapotranspiration showed an increasing trend, which may be related to the continuous increase in temperature in recent decades.
The CRcontribution rate of climate changeCC and HAhuman activities to streamflow changes accounted for 37.2% and 62.8%, 590 respectively, which was basically consistent with the results calculated by the new framework proposed in this study (the difference was 5.4%). According to the method introduced in Section 3.5, the changes in the streamflow caused by HAhuman activities in the LR Basin were separated, which mainly included the five sections of water consumption changes and the same amount of water depth 610 as the total dead storage capacity of the reservoir. Fig. 14 shows the CRcontribution rate of the five types of water withdrawals by HAhuman activities and the construction of the reservoirs to the streamflow changes in the LR Basin during the impacted period (from 2005 to 2015) compared to the natural period (from 1961 to 2004). Overall, the CRcontribution rate of HAhuman activities to streamflow changes was 59.91%, while that of climate changeCC was 40.09%. This result was also consistent with the results calculated in Section 4.3.1. Among them, the streamflow depth caused by the construction of the reservoir was reduced by -50.17 615 mm, which was also the factor that had the greatest impact on streamflow compared with other HAhuman activities, and its CRcontribution rate reached 58.0%, while the CRcontribution rate of the other five types of water withdrawal was relatively small.
The CRcontribution rates of domestic, irrigation, livestock, manufacturing, and mining water withdrawals were 1.32%, -0.35%, 0.12%, 0.79% and 0.03%, respectively, a total of 1.91%. In other words, the decrease in the streamflow in the LR Basin was mainly due to the impact of HAhuman activities, and most of it was caused by the construction of the reservoirs. 620

How does parameter uncertainty affect the quantitative results? 625
In this paper, we proposed a novel framework to quantify the CRcontribution rate of climate changeCC and HAhuman activities to streamflow changes considering the uncertainty of hydrological simulations. This is because the phenomenon of "equifinality for different parameters" in hydrological simulations greatly affects the quantification results. To preliminarily investigate the impact of model simulation uncertainty of the quantitative results, we selected the two simulation results with the largest NSEs in this study for analysis. The evaluation metrics and CRcontribution rate of climate changeCC and HAhuman activities are shown 630 in Table 7, which shows that both simulations can simulate the monthly streamflow at Yunjinghong station in the LR Basin accurately, and the two simulations have almost the same evaluation performance. However, the attribution analysis obtained from the two hydrological simulations showed completely different results. In the first simulation result, according to the method introduced in Section 3.4.1, the streamflow changes in the LR Basin were mainly caused by climate changeCC, but in the second hydrological simulation, the opposite conclusion was drawn, that is, HAhuman activities dominated. These were almost the same 635 hydrological simulation results but with opposite conclusions from the attribution analysis; this was one of the reasons why we must consider the uncertainty of the model parameters in the attribution analysis of climate changeCC and HAhuman activities using hydrological simulations. The results of Section 4.3.1 and related published studies (Han et al., 2019) in the LR Basin show that the streamflow changes in the LR Basin were mainly caused by HAhuman activities. (Notation: CC and HA represent the climate change and human activities, respectively; NSE and RE represent the Nash-Sutcliffe efficiency coefficient and the relative error, respectively) Table 8 shows the values of 9 highly sensitive parameters of the two simulation results and the streamflow values simulated by 645 the two simulations in the natural period and the impacted period. Table 8 and the calculation methods introduced in Section 3.4.1 show that the watershed streamflow reduction caused by climate changeCC calculated by the 1st and 2nd simulation results was -217.1 m 3 /s and -170.6 m 3 /s, respectively, which was the reason why they had opposing calculated attribution results. From the perspective of specific parameter values, the most sensitive parameter is ALPHA_BNK, which was the base flow alpha factor for bank storage (days) characterized by the bank storage recession curve. The difference between the two calibration results was not 650 large, and this parameter mainly controlled the baseflow process, having little effect on the average annual streamflow while the difference in CH_K2 in the two calibration results was larger, at 303.87 and 106.12. This parameter represented the effective hydraulic conductivity of the main channel alluvial layer, which meant that the larger the CH_K2 value is, the more likely the water in the main channel is lost to groundwater; accordingly, the streamflow production at the outlet of the watershed would decrease (Arnold et al., 2012b;Xu et al., 2016;Zhao et al., 2018a). This might also be one of the reasons that the first simulated 655 streamflow (1617 m 3 /s) was slightly smaller than the second one (1667.9 m 3 /s). The SFTMP parameter, which was the temperature when precipitation was converted into snowfall, returned values for the first simulation and the second simulation as 2.69°C and -0.11°C , respectively; this meant that in the first simulation, more liquid precipitation was converted into a solid state and less streamflow was formed, which also led to a smaller simulated streamflow in the first simulation. The SMTMP parameter, which was the snow melt base temperature, was -4.13°C in the first simulation result and 3.73°C in the second simulation result. 660 From basic physical knowledge, the SMTMP parameter in the second calibration result was more reasonable. Compared with other research results with similar terrain features in this study area, Debele et al. (2010) constructed the SWAT model in the high altitude area of the source of the Yellow River, China, and the SMTMP value obtained was 4°C . The difference between the two simulations was not large for the set of the other parameters (SOL_BD, GW_REVAP, CN2 and SOL_K), or the parameter that controlled the baseflow (ALPHA_BF) had little effect on the average streamflow of the basin. Based on the above, the second 665 simulation results were consistent with the calculation results of the new framework proposed in this study. Therefore, when we choose a hydrological simulation to analyze the attribution of climate changeCC and HAhuman activities to streamflow variations, we should clearly also consider the actual physical meaning and the uncertainties of the model parameters. Abbaspour et al. (2007); (Arnold et al., 2012a;Yang et al., 2017)  Simulated streamflow in the IP (m 3 /s) 1400.5 1497.3 (Notation: NP = "Natural period", IP = "Impacted period", R_, V_, and A_ represent multiplying, replacing, and adding the corresponding parameter values, respectively, in the process of calibrating the parameters.) In this study, 575 parameter combinations with good simulation results (NSE greater than 0.75) were selected, with a step size of 5%, it is proposed to reduce the influence of hydrological modeling uncertainty on the quantitative results by constructing the 675 posterior histogram distribution of the CR of CC and HAs to watershed streamflow change. However, it is undeniable that there are still unreasonable parameter combinations in the simulation results with high probability (167 times). For the LR basin, it is almost impossible to obtain the measured values of all 9 parameters with high sensitivity (Table 3). Therefore, in order to further explore the possible influence of unreasonable parameter values on the quantitative results, we selected two parameters related to snowmelt streamflow (SMTMP and SFTMP) to exclude unreasonable parameter combinations. According to the parameter value 680 ranges recommended by Abbaspour et al. (2007) and other related references (Arnold et al., 2012a;Yang et al., 2017), in this study, the reasonable value range of these two parameters is set to -5 to 5 ℃. After excluding parameter combinations outside this value range, we obtained 55 simulation results with relatively reasonable parameter values, and the quantization results obtained from this calculation are shown in Fig. 15. It can be seen from Fig. 15 that after excluding unreasonable parameter combinations, the calculated CR of CC in the LR Basin to the reduction of streamflow is 45-50% (with an average CR of 47.1%), and this result 685 is consistent with the results presented in Fig. 10 which derived from the novel framework proposed in our study. At the same time, it is also proved that although the calculation framework proposed in this study may contain unreasonable parameter combinations in obtaining the simulation results with the highest frequency, the calculation results are still highly accurate. In addition, for the research area where the measured values of related parameters can be obtained, the rationality and authenticity of the parameter values should be fully considered while selecting the parameter combination with higher NSE. 690 Fig. 15 Histogram of the number of simulations of the CR (with 5% steps) of climate change to streamflow reduction in the LR Basin at the annual scale and corresponding Nash-Sutcliffe Efficiency box plots after excluding the parameter combinations.

Land use/land cover change in the LR Basin from 1980 to 2015
In Section 4.4, the water withdrawals of domestic, irrigation, mining, livestock, and manufacturing, and in addition, dead 695 storage capacity of constructed reservoirs as well as the impact of HAhuman activities were separated; then the impacts of HAhuman activities on streamflow changes were separated. However, HAhuman activities also influenced the land use change on rainfall-runoff characteristics. Fig. 165 shows the land use in the LR Basin in1980, 2000, 2010 and 2015. Farmland was the largest land use in the upper LR Basin, while the lower reaches were dominated by forest. Due to the high-altitude terrain in the upper reaches, unused land and glaciers were mainly distributed in this area. Table 9 shows the areas of land use types in the LR 700 Basin in 1980Basin in , 1990Basin in , 2000Basin in , 2010Basin in and 2015 In general, the water area of the LR Basin showed a significant reduction from 1980 to 1990, which was possibly due to the decrease in the area of glaciers due to the increase in temperature from 1980to 1990). In contrast, the water area increased by nearly 30% from 2010 to 2015, which was mainly due to the construction of Nuozhadu hydropower station (with a total storage capacity of 22.7 km 3 ) within the basin. Table 9 Areas (km 2 ) of land use types in the LRLancang River Basin in 1980Basin in , 1990Basin in , 2000Basin in , 2010 Table 5 is second-level type which belong to Water) The area of farmland in the LR Basin showed a decreasing trend during 2000-2010 and 2010-2015, which is also the main reason for the reduction in the irrigation water consumption in the basin, which is consistent with the results shown in Fig. 13.
The areas of the cities all showed an increasing trend in the three periods of 1980-2000, 2000-2010 and 2010-2015 (by 29.8%, 14.1% and 27.1%, respectively), while the other three types of land use/land cover (i.e., forest, grassland, and unused land) did 710 not change significantly in the three periods. In summary, no significant changes were found from 1980 to 2015 in the forest and grassland of the LR Basin (accounting for 37.3% and 47.5% of the total area, respectively). Although the city area has undergone significant changes, it accounts for a very small total area of the basin very (0.09%). The change in the water area was mainly due to the construction of the reservoirs, so the method used in Section 4.3 to separate the contribution of HAhuman activities to the reduction in the streamflow in the LR Basin used is reasonable. 715

Comparison with results of other published studies
As analyzed above, there was no particularly significant change in the precipitation and potential evapotranspiration from 720 1961 to 2015 in the LR Basin. HAHuman activities mainly included the construction of reservoirs, resulting in changes in the streamflow. Attribution analysis results showed that the CRcontribution rate of HAhuman activities was 57.6%, and the corresponding climate changeCC was 42.4%. This result was basically consistent with Han et al. (2019), but the CRcontribution rate of HAhuman activities was smaller than the results of Han et al. research results (95%). This may be due to the following reasons: 725 1) The streamflow data of different time spans were used to obtain different break points. They used streamflow data from 1980 to 2014 to obtain the break point in 2008, and this study used data from 1961 to 2015 to identify the break point in 2005.
2) Different hydrological models were used. They used the coupled routing and excess storage (CREST) model with an NSE of 0.57, while the SWAT model used in this study had an NSE of 0.94.
3) Longer series of streamflow data and simulation data were used. 730 As indicated by Li et al. (2017a) and (Han et al., 2019), as the streamflow data series became longer in the impacted period, the impact of reservoir scheduling on the streamflow changes on an average scale for many years gradually decreased. Li et al. (2017a) selected Chiang Saen station, which was the nearest station to Yunjinghong station downstream of the LR Basin, for their research, and then they divided the streamflow series into three stages, the pre-impact period , the transition period (1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009) and the post-impact period (2010)(2011)(2012)(2013)(2014). They concluded that the construction of the reservoirs in the LR Basin led 735 to a decrease in the streamflow process during the flood period and an increase in the dry period, which was consistent with the results of our study (Section 4.3.2). Their results also showed that HAhuman activities contributed 61.88% to the streamflow reduction at Chiang Saen station, which was also close to the results of our study (57.4%).

Applicability and uncertainty of the proposed framework
A new quantitative framework for calculating the contribution rateCR of climate changeCC and human activitieHAs to 740 watershed streamflow variations was proposed in this study, and it was successfully applied to the Lancang RiverLR Basin with relatively accurate results. From our perspective, this method can effectively avoid reducequantify the influence of the phenomenon ofthe "equifinality for different parameters" that may exist in the use of hydrological simulation methods to quantify the CR of CC and HAs. At the same time, we also believe that this framework can be applied to other watersheds based on the following aspects. First, in this studythe section 4.4, we used tthe Budyko framework and the sectional water withdrawal data 745 within the basin were used to compare with the new framework. Second, the results of the comparison with published research on the Lancang River Basin (Han et al., 2019) also proved that the framework has good accuracy and applicability. Third, in the process of comparing with the new framework, we fully considered the impact of various HAhuman activities within the study area, including five types of water withdrawals (i.e., irrigation, livestock, living, mining, and manufacturing), the impact of reservoir storage and the land use/land cover change. Of course, due to the highly nonlinear relationship between the parameters 750 of the hydrological model, we suggest that readers ensure that the selected simulation results with NSEs greater than 0.75 are large enough when applying the novel framework in other research areas (this study had 500 simulations). It is undeniable that this method still has certain uncertainties and limitations when it is applied to other watersheds.Although this quantitative framework has been proved to have good applicability in the Lancang River Basin, it may also have certain limitations. First, if there are multiple break points in the annual streamflow sequence, then when selecting the unique break point, it is necessary to consider 755 the abrupt change points of the time series of other meteorological elements (precipitation, temperature, etc.) in the basin. At the same time, the impact of strong human activities (reservoir construction, large-scale water transfer project construction, etc.) on the abrupt change of streamflow in the basin should also be consideredsince there is only one break point in the mean annual streamflow sequence of the Yunjinghong station in the LR basin, there may be multiple break points in other study areas. If there exists more than one break points, the results of the break point inspection of the mean annual precipitation and potential 760 evapotranspiration in the study area, as well as the construction time of large reservoirs in the study area, should be considered (Dey and Mishra, 2017). Finally, a unique break point is selected to divide the research time series into a natural period and an impacted period, and then the quantitative framework proposed in this study can be applied. Second, because the SWAT model has good applicability at the Yunjinghong station in the LR Basin, it can meet the 500 best simulation requirements set by the framework proposed in this study, but the hydrological model may have different applicability in different research areas. 765 Therefore, the application of this framework in other research areas may have limitations, which need to be further verified. Third, because this study uses the parameter combinations obtained by the natural period to input the meteorological element data of the impacted period for calculation, this may also bring uncertainty to the calculation results, which is usually called "transferability" (Fu et al., 2018).
Although the new quantitative framework proposed in this study considers the uncertainties in hydrological simulations, the 770 framework is still based on traditional hydrological simulation methods to separate the contribution rateCR of climate changeCC to streamflow change, and then to deduce the contribution rateCR of human activitieHAs. Therefore, inevitably, there are still uncertainties in the calculation process. For example, the construction of large-scale reservoirs and changes in land use/land cover (urbanization, etc.) are important factors that alter the climatic state of a local region, specifically in that they change the temporal and spatial distribution characteristics of local regional hydrometeorological elements (Li et al., 2017c;Degu et al., 2011). This 775 change in meteorological elements was regarded as part of the impact of climate changeCC in this study; however, it was also caused by both HAhuman activities (reservoir construction) and CC. On the other hand, there are uncertainties in the division of the natural period and the impacted period in this study, which assumed that the impact of HAhuman activities on streamflow changes in the natural period was negligible; however, there were almost no periods within a watershed that were completely unaffected by HAhuman activities, and the impact of HAhuman activities on streamflow variations in the natural period was 780 ignored in these studies. In this study, there was also strong disturbance of HAs during the natural period (i.e., reservoir construction: Manwan and Dachaoshan) ( Table 1). In addition, our study selected the NSE and RE as the objective function to calibrate the SWAT model, which may also bring uncertainties in the quantitative results. As indicated by (Gupta et al., 2009) and Gupta and Kling (2011), using the NSE as an objective function to calibrate a hydrological model may tend to underestimate the peak streamflow. Although the CRcontribution rate in our study was calculated by the average streamflow over multiple years, it 785 still brought a given amount of uncertainty to the quantitative results. Therefore, follow-up research should strengthen the optimization of the objective function and benefit from field investigation of the actual meaning of the parameters. Since the impacts of climate changeCC and HAhuman activities on the hydrological processes of the watershed are complicated and interconnected, it is still a challenge to completely separate the impacts of climate changeCC and HAhuman activities on streamflow variations (Xin et al., 2019). Further consideration should be given to quantify the impact of specific HAhuman 790 35 activities, such as land use change and water withdrawal, and then to separate the impact of climate changeCC and HAhuman activities on streamflow changes as completely as possible.

Conclusions
In this study, we proposed a new framework that considered the uncertainties of model simulations to quantify the CRcontribution rate of climate changeCC and HAhuman activities to streamflow changes. This framework was developed based 795 on the posterior histogram frequency distribution (PHD) of the CRcontribution rate of climate changeCC and HAhuman activities.
Then, we selected the LR Lancang River (LR) Basin for the case study. Over the past three decades, after the construction of the Manwan Reservoir in 1987, six large reservoirs were constructed within the basin before 2014. The streamflow process in the watershed also has significant changes on multiyear average and monthly scales. The Mann-Kendall monotonic trend test and the Mann-Kendall break point test were used to test the trend and identify the break point of the annual streamflow data at 800 Yunjinghong station within the period of 1961 to 2015. Then, the available period was divided into the natural period (before the break point) and the impacted period (after the break point). Afterwards, the SWAT model and the SUFI-2 method were used to construct the Posterior Histogram Distribution (PHD) of the CRcontribution rate of climate changeCC and HAhuman activities.
Finally, the Budyko framework and the basin wide gridded monthly sectoral water use (GMSWU) data set were used to compare with the newly proposed framework. The main conclusions of this study are as follows: 805 1) The new proposed framework can be used to quantify the CRcontribution rate of climate changeCC and HAhuman activities in the LR Basin which can fully solve the local optimal solution for hydrological simulation parameters in current related studies.
The results of comparison using the Budyko framework and Gridded Monthly Sectoral Water Use (GMSWU) data set also showed that the new framework has high accuracy (the error range is within 6%).
2) The break point of the streamflow sequence during 1961-2015 at Yunjinghong station was identified in 2005. The streamflow 810 significantly decreased (~ -22%) after 2005 compared with that of the natural period (1961 -2004), which was mainly due to the construction of the Xiaowan Reservoir in October 2004. Significantly reduced streamflow in the flood period and significantly increased streamflow during the dry period also occurred, which was mainly due to the capacity adjustment of the constructed reservoirs. The trend test results also showed that from 1961 to 2015, the annual streamflow in the LR Basin showed a significant decreasing trend at the α = 0.01 significance level, precipitation showed a nonsignificant decreasing trend, and mean temperature 815 showed a significant increasing trend at the α = 0.01 significance level.
3) The quantification results calculated using the new proposed framework showed that, on an annual scale, compared with the natural period of 1961 -2004, the CRContribution Rate of Climate ChangeCC and HAHuman Activities (CR of CC and HAs) were 40 -45% (with an average CRcontribution rate of 42.6%) and 55 -60% (with an average CRcontribution rate of 57.4%), respectively. The CR of climate changeCC and HAhuman activities derived from the Budyko framework were 37.2% and 62.8%, 820 respectively, and the error between the two calculation results was 5.4%. The CR of HAhuman activities calculated using the GMSWU data and the reservoirs dead capacities was 58.0%, which also proved that the new proposed framework in this study can be used in the LR Basin. 4) Quantitative analysis results on a monthly scale in the LR Basin showed that, except for June and November, streamflow changes in other months were caused by HAhuman activities. Further analysis showed that the streamflow in June during the 825 impacted period decreased by 6.9 mm compared with that in the natural period, while the precipitation and potential evapotranspiration decreased and increased by 20.2 mm and 8.83 mm, respectively; the streamflow decreased by 5.34 mm in November, while the corresponding precipitation and potential evapotranspiration changed by -7.43 mm and 5.52 mm, respectively.
In summary, this study provides a new calculation framework that considers the uncertainty of hydrological simulations to 830 quantify the CRcontribution rate of climate changeCC and HAhuman activities to streamflow changes. The results of this case study also provide a reference for understanding the dominant factors of streamflow changes in the LRLancang River Basin and improving water resource management measures for the transboundary Lancang-Mekong River Basin. Of course, this new proposed framework also needs to be applied and verified in more research areas. In addition, this framework only considers the dual impacts of climate changeCC and HAhuman activities. However, in practical applications, water resource decision makers 835 are more willing to understand the specific impacts of HAhuman activities such as irrigation water and land use changes. Therefore, in future research, efforts should be made to expand the framework to quantify the CRcontribution rates of individual items of climate changeCC and HAhuman activities.